I have started using AI to help me write a set of custom ComfyUI nodes that builds upon, expands and customizes what I call scaffolding using what I call environmental priors. I won’t go into the mechanics of that process here, but I will share some of the outputs that are helping me zero in on the features and fine-tune the settings that seem to work best.
Obviously, as experiments, some of these images will be better than others, and some may be considered as failures, but each has taught me something. Consider this gallery as a running lab journal – some will use the same or similar environmental prior and luminance map with slightly different scaffold settings, while others will be completely different. I will try to remember to section-off those that use a different prompt so that these images don’t become too confusing a mix of differences and lose their value as “lab notes.”
